Chest radiography is one of the most common types of diagnostic radiology exams, which is critical for screening and diagnosis of many different thoracic diseases. Specialized algorithms have been ...developed to detect several specific pathologies such as lung nodules or lung cancer. However, accurately detecting the presence of multiple diseases from chest X-rays (CXRs) is still a challenging task. This paper presents a supervised multi-label classification framework based on deep convolutional neural networks (CNNs) for predicting the presence of 14 common thoracic diseases and observations. We tackle this problem by training state-of-the-art CNNs that exploit hierarchical dependencies among abnormality labels. We also propose to use the label smoothing technique for a better handling of uncertain samples, which occupy a significant portion of almost every CXR dataset. Our model is trained on over 200,000 CXRs of the recently released CheXpert dataset and achieves a mean area under the curve (AUC) of 0.940 in predicting 5 selected pathologies from the validation set. This is the highest AUC score yet reported to date. The proposed method is also evaluated on the independent test set of the CheXpert competition, which is composed of 500 CXR studies annotated by a panel of 5 experienced radiologists. The performance is on average better than 2.6 out of 3 other individual radiologists with a mean AUC of 0.930, which ranks first on the CheXpert leaderboard at the time of writing this paper.
Abstract
Most of the existing chest X-ray datasets include labels from a list of findings without specifying their locations on the radiographs. This limits the development of machine learning ...algorithms for the detection and localization of chest abnormalities. In this work, we describe a dataset of more than 100,000 chest X-ray scans that were retrospectively collected from two major hospitals in Vietnam. Out of this raw data, we release 18,000 images that were manually annotated by a total of 17 experienced radiologists with 22 local labels of rectangles surrounding abnormalities and 6 global labels of suspected diseases. The released dataset is divided into a training set of 15,000 and a test set of 3,000. Each scan in the training set was independently labeled by 3 radiologists, while each scan in the test set was labeled by the consensus of 5 radiologists. We designed and built a labeling platform for DICOM images to facilitate these annotation procedures. All images are made publicly available in DICOM format along with the labels of both the training set and the test set.
Selective serotonin reuptake inhibitors (SSRIs) are among the popular drugs for treating depression and mental disorders. Membrane fluidity has previously been considered as the main factor in ...modulating the membrane partitioning of SSRIs, while other biophysical properties, such as the acyl chain order and area per lipid, were often neglected. Varying the lipid membrane composition and temperature can significantly modify the physical phase and, in turn, affect its fluidity, acyl chain order and area per lipid. Here, we investigate the role of membrane fluidity, acyl chain order and area per lipid in the partitioning of two SSRIs, paroxetine (PAX) and sertraline (SER). The model membranes were either POPC : SM (1 : 1 mol ratio) or POPC : SM : Chol (1 : 1 : 1 mol ratio) and studied in the temperature range of 25-45 °C. The order parameters and area per lipid in the two lipid mixtures were calculated using molecular dynamics simulations. The membrane partitioning of PAX and SER was determined
second derivative spectrophotometry. In a lower temperature range (25-32 °C), membrane fluidity favors the SSRI partitioning into L
/L
POPC:SM:Chol. In a higher temperature range (37-45 °C), the interplay between membrane fluidity, acyl chain order and area per lipid favors drug partitioning into L
POPC:SM. The findings offer indication for the inconsistent distribution of SSRIs in tissues as well as the possible interaction of SSRIs with lipid domains and membrane-bound proteins.
The chemical composition and larvicidal activity of essential oils from the leaves and rhizomes of Zingiber collinsii Mood & Theilade (Zingiberaceae) were reported. The main compounds in the leaf oil ...were α-pinene (25.6%), β-caryophyllene (16.8%), β-pinene (16.1%) and bicyclogermacrene (6.9%) while the rhizome oil consist mainly of camphene (22.5%), β-pinene (16.3%), α-pinene (9.0%) and humulene oxide II (9.0%). The rhizome oil demonstrated larvicidal effects towards fourth instant larvae of mosquito vectors. The highest mortality (100%) was observed at 24 h exposure against Aedes albopictus (concentration 100 μg/mL) and 48 h (concentration of 50 and 100 μg/mL), while the highest mortality (100%) was observed for Culex quinquefasciatus at 24 h and 48 h at concentration of 100 μg/mL. The 24 h mosquito larvicidal activity of the rhizome oil against Ae. albopictus were LC50 = 25.51 μg/mL; LC90 = 40.22 μg/mL and towards Cx. quinquefasciatus with LC50 = 50.11 μg/mL and LC90 = 71.53 μg/mL). However, the 48 h larvicidal activity were LC50 = 20.03 μg/mL and LC90 = 24.51 μg/mL (Ae. albopictus), as well as LC50 = 36.18 μg/mL and LC90 = 55.11 μg/mL (Cx. quinquefasciatus). On the other hand, no appreciable mortality and larvicidal activity was observed for the leaf oil. The larvicidal activity of the essential oils of Z. collinsii was being reported for the first time.
The expansion of fintech credit around the world is challenging the global banking system. This study investigates the interrelationships between the development of fintech credit and the efficiency ...of banking systems in 80 countries from 2013 to 2017. The findings indicate a two-way relationship between them. More specifically, a negative relationship between bank efficiency and fintech credit implies that fintech credit is more developed in countries with less efficient banking systems. Meanwhile, a positive impact of fintech credit on the efficiency of banking systems suggests that fintech credit may serve as a wake-up call to the banking system. Therefore, fintech credit should be encouraged by the authorities around the world.
The Covid-19 pandemic's economic effect led to tighter credit standards and a decline in the market for many types of loans. With a rich database of 1,231 banks in 90 countries from 2018Q1 to 2021Q4, ...we conducted a timely, broad-based international study to investigate whether non-interest activities, serving as a shock absorber, can promote bank performance before and during the Covid−19 pandemic. When using a dynamic panel data model with a system GMM estimator, our findings indicate that banks should be encouraged to diversify their income sources to reduce the adverse effects of the shock. With comparative analysis, we also found heterogeneous effects of income diversification on bank performance by its components, in pre-Covid−19 and during-Covid−19 periods, in both developed and developing countries. This study implies that bank managers should diversify income sources, especially fee-based services, trading activities, and foreign currency, to foster financial performance and stability during exogenous shocks.
This study investigates the relationship between market power and bank profitability, and the impacts of CEOs’ personality traits, in Vietnam from 2007 to 2020. The analysis of CEOs’ signatures is ...used to determine their characteristics. The findings support the quiet-life hypothesis, which suggests that the negative relationship between market power and bank profitability may depend on CEOs’ characteristics. More specifically, the results show that conscientious CEOs with market power tend to reduce bank profitability, and this effect is more pronounced for foreign-owned banks. Therefore, our findings have critical implications for bank management.
Knowledge of thermodynamics of lipid membrane partitioning of amphiphilic drugs as well as their binding site within the membrane are of great relevance not only for understanding the drugs' ...pharmacology but also for the development and optimization of more potent drugs. In this study, the interaction between two representatives of selective serotonin reuptake inhibitors, including paroxetine and sertraline, and large unilamellar vesicles (LUVs) composed of 1,2-dioleoyl-
-3-phosphocholine (DOPC) was investigated by second derivative spectrophotometry and Fourier transform infrared spectroscopy (FTIR) to determine the driving force of the drug partitioning across lipid membranes. It was found that temperature increase from 25 to 42 °C greatly enhanced the partitioning of paroxetine and sertraline into DOPC LUVs, and sertraline intercalated into the lipid vesicles to a greater extent than paroxetine in the temperature range examined. The partitioning of both drugs into DOPC LUVs was a spontaneous, endothermic and entropy-driven process. FTIR measurements suggested that sertraline could penetrate deeply into the acyl tails of DOPC LUVs as shown by the considerable shifts in the lipid's CH
and Cdouble bond, length as m-dashO stretching modes induced by the drug. Paroxetine, however, could reside closer to the head groups of the lipid since its presence caused a larger shift in the PO
bands of DOPC LUVs. The findings reported here provide valuable insights into the influence of small molecules' chemical structure on their molecular interaction with the lipid bilayer namely their possible binding sites within the lipid bilayer and their thermodynamics profiles of partitioning, which could benefit rational drug design and drug delivery systems.
This study explains the differences and variances in the efficiency scores of the Vietnamese banking sector retrieved from 27 studies published in refereed academic journals under the framework of ...meta-regression analysis. These scores are mainly based on frontier efficiency measurements, which essentially are Data Envelopment Analysis (DEA) and Stochastic Frontier Analysis (SFA) for Vietnamese banks over the period of 2007-2019. The meta-regression is estimated by using truncated regression to obtain bias-corrected scores. Our findings suggest that only the year of publication is positively correlated with efficiency, whilst the opposite is true for the data type, and sample size.
Digital credit has gained much attention from academic researchers, practitioners, and policymakers worldwide. This study empirically evaluates the determinants of digital credit using cross-country ...data from 2013 to 2019. The conventional ordinary least square regression with fixed effects estimator is used to investigate the factors affecting the growth of digital credit. Our study highlights that the regulatory frameworks of anti-money laundering and terrorist financing, the economy’s innovative capacity, and financial development are significant factors affecting the development of digital credit, especially fintech credit. However, the findings indicate that only the innovation capacity is more critical to the expansion of bigtech credit. Nonetheless, our results provide some important implications for market participants and the authorities in promoting digital credit. Accordingly, this study contributes to the literature on the growth of digital credit when considering the critical roles of money laundering and terrorist financing frameworks and innovation capacity.
PDF
